Scientists on Wednesday announced an important advance in the long quest to harness nuclear fusion, a field that has sparked dreams of clean and limitless energy.

Fusion, the process that powers the Sun and other stars, entails forging the nuclei of atoms to release energy, as opposed to splitting them, which is fission -- the principle behind the atomic bomb and nuclear power.

Decades of work in fusion have run up against a giant hurdle -- the energy yield from the reaction has been dwarfed by the vast amounts of energy needed to trigger the process.

But, in lab experiments described by scientists in the United States, major progress has been made in overcoming this obstacle.

Reporting in the journal Nature, researchers said they were the first to tease more energy out of a fusion reaction than had been absorbed by the fuel used to spark it.

They fixed 192 laser beams onto a spot narrower than the width of a human hair to generate enough energy to compress a tiny fuel-containing capsule to a 35th of its original size.

Lasting less than a billionth of a second, the reaction put out the equivalent of the energy stored in two AA batteries (some 17,000 joules) in their latest experiment in November 2013.

Though "modest", according to the team, the output was higher than the estimated 9,000-12,000 joules of energy taken up by the fuel.

"This is closer than anyone has gotten before" to generating viable fusion energy, the study's chief author Omar Hurricane of the US government-run National Ignition Facility (NIF) in California said.

The yield was 10 times greater than previously achieved.

There are qualifiers, though.

It was not a sustained reaction, an eagerly sought moment called ignition.

And it still does not answer the efficiency challenge of releasing more fusion energy than is consumed overall.

In this case, the lasers put out about 1.9 million joules of energy -- the equivalent energy in a small car battery -- of which only 9,000-12,000 joules were absorbed by the fuel.

"Only something like one percent of the energy that we put in from the laser ends up in the fuel right now, maybe even less," said co-author Debbie Callahan.

"There is a lot of room for improvement."

The method needs to be refined and the yield boosted 100 times "before we get to the point of ignition," Hurricane added.

"We can't honestly tell you when we will have ignition. We are working like mad to go that direction. Our theoretical understanding says if we keep pushing in this direction, we have a chance."

Ignition also requires self-propagation, in which the first fused particles cause the heat and pressure to build even further, thus creating more particles, and so on, to boost the energy yield.

The latest experiments at the NIF, one in September last year and the other in November, were the first to yield evidence of particles "leaving some energy behind", Hurricane said.

- Recreating conditions at Sun's core -

Nuclear fusion is a process by which the nuclei of deuterium and tritium, both isotopes obtained from hydrogen, are fused together to create heavier particles.

In theory, energy generated through fusion would leave no dangerous waste or pollute the atmosphere. And the fuel is found in abundance in seawater, which covers more than two-thirds of the planet.

The process requires extreme temperatures and pressure equivalent to those found on our Sun and other active stars.

To achieve this, Hurricane and his team shot their lasers at a gold cylinder two millimetres (0.08 inches) in diameter that was coated on the inside with a frozen layer of the deuterium-tritium fuel.

The light entered through holes on one end and re-focused in X-rays that blasted off the cylinder's outer shell and caused the remainder to implode on a scale likened to shrinking a basketball to the size of a pea.

The process generated pressure 150 billion times stronger than that exerted by Earth's atmosphere and a density 2.5 to three times greater than the core of the Sun, the scientists said.

In a comment on the findings, fusion researcher Mark Herrmann of the Pulsed Power Sciences Center in Albuquerque said they were "a significant step forward".

There are two research tracks in fusion energy.

One, like the NIF, uses lasers, while the other uses large-scale magnetic fields -- the approach adopted by the International Thermonuclear Experimental Reactor, a 15 billion euro ($19.5 billion) project in southern France set to become operational in 2019.